Context-manager progress bar for distributed tasks

Currently the only way to get a progress bar for distributed tasks is to use `distributed.diagnostics.progressbar.progress(futures: Futures)`. However, to use this, we need to have a list of futures. When using some of the higher level distributed constructs such as `dask.dataframe.DataFrame.compute()`, we don't explicitly get access to the futures. We can technically access them, but it gets a bit ugly. Instead I wonder if we could have a context manager progress bar that works like the local scheduler one. I envisage it would work like this

```python
from dask.distributed import Client
import dask.dataframe as dd
from distributed.diagnostics.progressbar import Progress

client = Client()
with Progress():
df = dd.read_csv('2014-*.csv').apply(...).compute()
```